[Phil_M]

CO2 doesn't act like a kind of invisible water, it's much more buoyant, even if heavier than air. I reckon it would dissipate quickly if given a stir.

Liquids don't float freely out of containers - gravity keeps them in place at the bottom of the flask. So they're very different in this context, despite both being fluids.

Wait, what? No, CO2 doesn't have the mass of water. But it DOES act like an invisible water...or else it wouldn't be a fluid. Now, have you or have you not seen CO2 being poured over a candle, like water, to extinguish the candle?

Or stuck your nose anywhere near a sink after you've dumped a keg in it? It takes a good while for CO2 to move, or a really strong air currant.

[a10t2]

Hopefully, you are not basing your assertion on Kai's work.  A lot his work is so horribly flawed that it has me casting doubt on the accuracy of all of it.

Along with my own, Jamil Z's, Chris White's, etc. The correlation seems pretty well-established, even if the mechanism is ambiguous.

[charles1968]

CO2 doesn't act like a kind of invisible water, it's much more buoyant, even if heavier than air. I reckon it would dissipate quickly if given a stir.

Liquids don't float freely out of containers - gravity keeps them in place at the bottom of the flask. So they're very different in this context, despite both being fluids.

Wait, what? No, CO2 doesn't have the mass of water. But it DOES act like an invisible water...or else it wouldn't be a fluid. Now, have you or have you not seen CO2 being poured over a candle, like water, to extinguish the candle?

Or stuck your nose anywhere near a sink after you've dumped a keg in it? It takes a good while for CO2 to move, or a really strong air currant.

Gases spread out to fill the container they occupy, whereas liquids maintain a fixed volume and are pulled down by gravity. They are different states of matter and behave in fundamentally different ways. You're right that CO2 is a fluid, but don't confuse that with a liquid. You can pour CO2 over a candle to extinguish it but the CO2  will spread out and dissolve in the air, whereas a liquid would fall to the floor.

[Phil_M]

But at what rate? Diffusion is not an instantaneous process. Couple that with an active CO2 source, and you have a situation quite different from that described by the ideal gas law.

[charles1968]

If two gases are moving, they will mix very quickly, like milk being stirred into coffee. If they're motionless then they'll mix by diffusion only and yes that's a lot slower. A flask on a stir plate doesn't contain motionless gas.

[Phil_M]

I've conducted a quick-an-dirty experiment.

I used my stir plate set to 11 and a 3" stir bar, here's a picture of it stirring 1L of water. Figured this was about as much of a vortex as anyone was likely to get.



As I don't have any matches handy, I used a hunting wind indicator. Uses talcum or some other sort of superfine powder to show air currents.

I sampled the air currents in these ways:

Inside the flask, no liquid. Powder rapidly rose out of the flask, noticeable swirling air both in and out of the flask.

Inside the flask, no liquid, with foam stopper. Powder did not appear to escape stopper. (This is an area where the powder isn't a good analog for a gas. But, then neither is smoke.) I used a laser level to watch for particles, didn't see any.

Inside the flask, with liquid, no foam stopper. No noticeable vortex outside the glass, slight vortex inside. The powder that did escape the flask simply wafted out, very much like smoke. There where no signs of circulation.

I also tried no liquid with a foam stopper, firing the powder above the stopper to try and see any air currents. There did not appear to be any, and at this point I didn't see the point in repeating these tests with a loosely crimped piece of foil.

I standby my initial assumption that with any sort of cover diffusion better describes the exchange of gasses between the flask and the outside air.

[narvin]

Lower *partial* pressure, though. As CO₂ displaces oxygen, diffusion acts to draw more in. If there's no significant O₂ in the headspace, my question is the same as before: what's the mechanism for a starter with an airlock growing less yeast?

Hopefully, you are not basing your assertion on Kai's work.  A lot his work is so horribly flawed that it has me casting doubt on the accuracy of all of it.

I think it's worth mentioning that his "work" of explaining the science and art of homebrewing also includes a lot of references and citations from brewing texts.  The blog is where the one-off experiments are.  Hopefully, you understand the difference.

[charles1968]

I've conducted a quick-an-dirty experiment.

I used my stir plate set to 11 and a 3" stir bar, here's a picture of it stirring 1L of water. Figured this was about as much of a vortex as anyone was likely to get.



As I don't have any matches handy, I used a hunting wind indicator. Uses talcum or some other sort of superfine powder to show air currents.

I sampled the air currents in these ways:

Inside the flask, no liquid. Powder rapidly rose out of the flask, noticeable swirling air both in and out of the flask.

Inside the flask, no liquid, with foam stopper. Powder did not appear to escape stopper. (This is an area where the powder isn't a good analog for a gas. But, then neither is smoke.) I used a laser level to watch for particles, didn't see any.

Inside the flask, with liquid, no foam stopper. No noticeable vortex outside the glass, slight vortex inside. The powder that did escape the flask simply wafted out, very much like smoke. There where no signs of circulation.

I also tried no liquid with a foam stopper, firing the powder above the stopper to try and see any air currents. There did not appear to be any, and at this point I didn't see the point in repeating these tests with a loosely crimped piece of foil.

I standby my initial assumption that with any sort of cover diffusion better describes the exchange of gasses between the flask and the outside air.

It's interesting you saw a slight vortex in the air space - that shows the gases are moving. But to demonstrate that diffusion is the main source of gas exchange with the outside rather than the vortex, you would need to compare a no-stir flask (still air; diffusion only) to a flask with stirring (vortex+ diffusion). I still think the stirred flask will draw in oxygen, even with a cover. Only an airtight lid would stop oxygen getting in.

[S. cerevisiae]

I still think the stirred flask will draw in oxygen, even with a cover. Only an airtight lid would stop oxygen getting in.

I did too until Martin got me thinking about the problem in greater detail (darn you Martin    ).   Inside of the flask, we have a situation where the gas is at higher pressure than outside of the flask.  According to Boyle's law the volume of a gas is inversely proportional to its pressure. As the pressure of a gas increases, so does its density, and the space between the molecules decreases (not to mention that CO₂ is also a denser molecule than O₂). In order for O₂ molecules to diffuse into a flask, they have to overcome a low pressure to high pressure gradient and find empty spaces between CO₂ molecules that are evacuating the flask.  Any O₂ that manages to make it past that barrier is insignificant.

[Phil_M]

You ask us to provide numbers, yet provide none yourself. By that I take it that none of us are aero engineers. I'm drawing on my chemistry and physics classes, and while I don't have numbers, I have enough experience to know that it's unlikely that a significant amount of oxygen is entering the flask. I even performed a test that backed up those initial conclusions. Mark also makes a good point about the pressures being different, which also seems to back up my simple experiment.

I could be wrong, I admit that. Meanwhile you berate us for not providing numbers, while also not providing numbers. Unless someone shows some other test, mechanism, law, or theorem to explain how the oxygen gets in the flask, I stand by my statement.

And as a senior technician and engineering student, yes, theories do get strung out together. Then you build your formulas to model the situation, then plug in the numbers. Right now I'm not seeing any mechanism for significant levels of oxygen to enter the flask, so there's no need to run the numbers. It's like looking at my desk, seeing a pencil on it, then deciding I don't have to balance the forces to see if the pencil will accelerate or not.

We have plenty of theorems that show why oxygen shouldn't enter the flask, please show us one that shows why it should.

[a10t2]

We have plenty of theorems that show why oxygen shouldn't enter the flask, please show us one that shows why it should.

Can you elaborate on how you think diffusion falls short?

[charles1968]

You ask us to provide numbers, yet provide none yourself.

Actually there's a lot of data in the braukaiser links techbrau posted.

[charles1968]

I still think the stirred flask will draw in oxygen, even with a cover. Only an airtight lid would stop oxygen getting in.

I did too until Martin got me thinking about the problem in greater detail (darn you Martin    ).   Inside of the flask, we have a situation where the gas is at higher pressure than outside of the flask.  According to Boyle's law the volume of a gas is inversely proportional to its pressure. As the pressure of a gas increases, so does its density, and the space between the molecules decreases (not to mention that CO₂ is also a denser molecule than O₂). In order for O₂ molecules to diffuse into a flask, they have to overcome a low pressure to high pressure gradient and find empty spaces between CO₂ molecules that are evacuating the flask.  Any O₂ that manages to make it past that barrier is insignificant.

Numeric simulation or it doesn't count.

This is what I was talking about w.r.t. hand waving in the other thread (I don't mean to imply that you're the only one guilty of this, either, but peel away the jargon and it's what many of your wall of text arguments/pontifications boil down to). Engineers don't get to design bridges by stringing together cherry picked physical theories into narratives without paying any mind to the actual numbers. The fact that the fermenting wort is a carbon dioxide source does imply that there must be a pressure gradient between the inside and outside of the flask, but I have a hard time believing that the magnitude of this gradient is anywhere close to great enough to do what you're describing.

Say what you will about Kai, his experiments were a hell of a lot more solid than most in homebrewing. I agree that there are factors which he did not take into consideration (e.g. did not measure the initial viability, did not measure the initial sterol and UFA reserves of the yeast going in to the ferment, did not measure the initial DO of the wort in each flask, tested on only one yeast strain...) and that he did not perform multiple trials for each configuration, but he did do a string of 3 different experiments concerning access to air/stirring speed and replicated the same trend in each, namely that more air and/or a higher stirring speed produced more growth. This is not enough to convince me completely, but it's enough to raise my eyebrows.

http://braukaiser.com/blog/blog/2013/03/19/access-to-air-and-its-effect-on-yeast-growth-in-starters/
http://braukaiser.com/blog/blog/2013/03/25/stir-speed-and-yeast-growth/
http://braukaiser.com/blog/blog/2013/05/19/yeast-growth-and-the-question-of-quality-vs-quantity/

Assuming that dissolved oxygen is the limiting nutrient for yeast growth AND viability in terms of built up sterol and UFA reserves, if more experiments were conducted and controlled more tightly - especially regarding the initial levels of dissolved oxygen in each test wort - demonstrating that a stirred, foil covered starter consistently grew more cells with higher viability (e.g. storing the cultures for a month and then staining them to do viability counts) than an unstirred culture with an airlock, then we must conclude that stir plates do in fact draw oxygen in to the flask.

All that said, at this point I'm considering ditching my stir plate and using an aquarium pump with an in line HEPA filter to continuously aerate my starters.

Interesting links. The general trend of faster spinning -->> more yeast does suggest stir plates draw in more O2, but the lower growth at highest speed is also consistent with Mark's ideas about shear stress.

The weight loss graphs in the third link are also v interesting.

[Phil_M]

Is there some other mechanism besides oxygen that can lead to this perceived cell growth? (I say perceived, because as I understand it, methylene blue isn't the greatest indicator of yeast health/vitality...)

Does the shear stress cause buds to break off early, leaving the parent cell with more energy reserves that are then spent making another bud?

[Phil_M]

Maybe, just don't forget about Occam's Razor 

I have yet to see any convincing evidence whatsoever that shear stress at the home brew level is not just a bogeyman. And yes Mark, I googled "magnetic stirrer" and "shear stress."

Yes, I'm familiar with it, nor have I forgotten it. If everything points to insignificant amounts of oxygen entering the flask, wouldn't some other mechanism involve fewer assumptions? At this point we seem to be relying on some unknown mode of O2 entry into the wort. Some other mechanical affects seem the next most likely source of increased growth.